Predictors of Post-traumatic stress symptomology in parents of infants with Congenital Heart Disease post-surgery and after four months

Abstract

Purpose:

To identify predictors of post-traumatic stress symptomology among parents of infants with complex congenital heart defects at hospital discharge and after four months.

Design & Methods:

A secondary analysis utilizing data from a larger RCT performed in three pediatric cardiac centers in North America. Analysis included 158 parent-infant dyads. Generalized Linear Modeling was used to identify predictors of parental post-traumatic symptomology at hospital discharge, and after four months. Considered predictors included demographics/SES, illness, and psychosocial parameters.

Results:

At discharge, parenting stress, education, and infant’s medication number were linked to post-traumatic stress symptomology severity; Parenting stress, education, insurance type, and medications number predicted number of symptoms; Tube-assisted feeding predicted PTSD. At four months, parenting stress, ethnicity, and number of ED visits predicted PTSS severity; Parenting stress, ethnicity, and cardiologist visits predicted number of symptoms; Parenting stress, single ventricle physiology, and number of children predicted PTSD.

Conclusions & Practical Implications:

Parental psychosocial factors, additionally to illness and sociodemographic indicators, can potentially risk parents to experience PTSS/PTSD. Nursing and other healthcare professionals can participate in early screening of such factors to determine familial risk.

Introduction

Congenital heart disease (CHD) affects 12 million children worldwide. In the United States, one-quarter of all infants born with CHD require invasive interventions to survive past the neonatal period. Whereas the surgical and medical treatments have tremendously improved in the last decades with increasing survival rates of above 80%, the disease still poses a threat on the lives of infants with more complex malformations (e.g. Hypoplastic Left Heart Syndrome, Tetralogy of Fallot, Transposition of the Great Arteries) (Zimmerman et al., 2020). These infants require corrective or palliative surgeries during the early weeks of life, followed by lengthy hospitalizations in the cardiac critical care units (CICU), and place extensive caretaking burden and distress on their parents post discharge and over the following months (Woolf-King et al., 2017).

Increased recognition is being given to the parental psychological distress in the CHD population, and the related overall health and wellness of parents. Parents of infants with CHD have been demonstrating increased parenting stress, anxiety, social dysfunction, and decreased quality of life across studies (Bishop et al., 2019; Woolf-King et al., 2017). According to various family stress frameworks and to the related body of literature, the parental stress is multidimensional and evokes by a complex interaction of factors related to the child, the parent, and the environment (Golfenshtein et al., 2015; Hearps et al., 2014; Sarajuuri et al., 2012). Specifically with regard to the CHD population, such factors include, for example, the child’s temperamental characteristics, illness severity parameters, parental perceptions of their competence and parenting role alteration, social support, partner relationships and other socioeconomic factors (Lisanti, 2018; Torowicz et al., 2010). These psychosocial outcomes are often associated with reduced parental quality of life and well-being in comparison with population norms (Bishop et al., 2019; Ernst et al., 2018; Hearps et al., 2014; Landolt et al., 2011).

Recent studies show that parents in the CHD population may also suffer from post-traumatic stress symptoms (PTSS) and meet diagnostic criteria for post-traumatic stress disorder (PTSD) (Landolt et al., 2011; McWhorter et al., 2021; Medoff Cooper et al., 2020), resulting from their experience of the cardiac surgery and the critical care period as traumatic events. The scant research in this population reported on symptom prevalence rates ranging from 11% to 68% ((Landolt et al., 2003; Medoff Cooper et al., 2020; Woolf et al., 2016), when symptoms can include distressing and intrusive thoughts, avoidance, and hyperarousal. A significant proportion of those parents also meet the previous DSM diagnostic criteria (American Psychiatric Association, 2013) for PTSD (Helfricht et al., 2008; Landolt et al., 2011;; Medoff Cooper et al., 2020). Studies in other pediatric populations have investigated parental PTSS and PTSD much more in depth (Boyer et al., 2000; Chardon et al., 2021; Kazak, Alderfer, Rourke, et al., 2004). In such populations, parents suffering from post-traumatic stress symptomology demonstrated poor parental mental health outcomes, decreased quality of life, strains in the parental role, and greater utilization of pediatric health services (Cook et al., 2018; Ernst et al., 2018; Helfricht et al., 2008; Landolt et al., 2003; Parfitt & Ayers, 2009; Raphael et al., 2010; Thompson et al., 2017). Therefore, it is critically important to recognize the factors that play a role in the parental post-traumatic stress in this population, especially in mothers, who appear to be more at-risk for adverse mental health outcomes in the critical infancy period. Yet no studies have examined such factors in the CHD population.

Several models together with the available research provide a framework for identifying potential risk factors of parental PTSS/PTSD following pediatric medical traumatic events (Kazak et al., 2006; Lisanti, 2018;Lisanti et al., 2017; Muscara et al., 2015). These risk factors are broadly categorized into demographic factors, illness-related factors, and psychosocial factors. The dynamics of these factors’ impact is also acknowledged, as PTSS and PTSD develop and change over time and within the context of the illness phases (Muscara et al., 2018; Woolf et al., 2016). The aim of the current study was to identify predictors of post-traumatic stress symptomology among parents of infants with complex congenital heart defects at hospital discharge and after four months.

Methods

This descriptive secondary cross sectional study utilized data from a previously described randomized clinical trial conducted from 2012–2017 (Trial Registration Blinded), in which parents in the intervention group received a telemedicine intervention supporting parents caring for their infant’s after discharge home post-surgery, compared to parents in the control group. Subjects were recruited from the cardiac intensive care units of three large, university-affiliated pediatric cardiac centers in the American Northeastern region. The study was approved by the institutional review boards. The main study included 219 parent-infant dyads of infants with complex CHD who had undergone cardiac surgery in the first three weeks of life. Eligible infants were at least 37 weeks gestation, weighed 2,500 grams at birth, and had a CHD with a Risk Adjustment in Congenital Heart Surgery (RACHS-1) Category of 2 or greater. Exclusion criteria included parents who did not speak or read English, were less than 18 years of age, or had infants with genetic disorders and other syndromes (except for non-syndromic appearing DiGeorge syndrome), cardiomyopathy and/or those awaiting heart transplant. Participants signed written informed consent. Data were collected prior to the infant’s hospital discharge, and at 4 months post-discharge follow-up. For a detailed description of the main study, see Medoff Cooper et al., 2020. Sample for the current study included 158 parent-infant dyads from both the intervention and the control groups who had complete PTSD data, as groups did not significantly differ on any of the current study’s variables of interest, including the PTSS/PTSD outcomes. No significant differences in participants’ characteristics were found between the main study’s sample and the current sample.

Study variables and Measures:

Parental Post-Traumatic Stress:

Parents completed the Posttraumatic Diagnostic Scale (PTDS) (Foa et al., 1997) within two weeks from discharge, and at the four months follow-up. The PTDS is a self-report measure of PTSS and PTSD designed for individuals who were exposed to high-magnitude stressors. Forty-nine items measure the number, type and severity of symptoms, corresponding to the 17 listed in the DSM-IV criteria (American Psychiatric Association, 2000) over five continuous subscales (Number of Symptoms, Re-experiencing, Avoidance, Hyper arousal, Total symptom Severity). Higher scores represent worse symptoms. Additional subscale allows to determine whether the individual adheres the PTSD diagnostic criteria according to the DSM (yes/no score). The PTDS has been validated in various clinical and research setting and demonstrates high internal consistency (α=0.92) and test retest reliability (r=0.74) (Foa et al., 1997). Parents were asked to complete the PTDS items as they related to their child’s cardiac condition as the traumatic event. The main outcomes of interest in the current study were the Number of Symptoms subscale, Total Symptom Severity score, and PTSD diagnosis, at discharge and at four-month follow-up.

Potential predictors considered for analysis

Demographic Factors:

The following parental and infant information considered as potential contributors to the parents’ PTSS/PTSD were collected via self-parental reporting and included infant’s and parent’s sex, parental age, parental race/ethnicity, educational level, insurance type, income, and number of children in house.

Illness-related factors:

The child’s factors related to the cardiac condition considered as potential predictors of parental PTSS/PTSD included birthweight, timing of cardiac diagnosis (prenatal vs. postnatal), and various illness severity indicators such as RACHS-1 score (Al-Radi et al., 2007), cardiac physiology (single vs. two ventricle physiology), hospital length of stay in days, number of medications, number of unplanned hospital readmissions and cardiologist visits over the study period, feeding route (oral vs. tube assisted), and growth failure defined as infant WHO weight-for-age z-scores ≤2 (World Health Organization, 2006).

Psychosocial Factors:

The perceived parenting stress and additional parenting role aspects, the perceived social networks and the quality of close relationships were considered as potential predictors of parental PTSS/PTSD.

Parenting Stress was measured at discharge and at follow-up via the Parenting Stress Index (PSI, long form (Abidin, 2012). The PSI assesses 13 stress-evoking aspects in the parenting system, which are divided into Parent and Child Domains. The Child Domain includes a child’s temperament and behavioral characteristics, as well as parental expectations with regard to the child. The Parent Domain includes aspects related to the parent’s characteristics, expectations, and feelings about the parenting role and the parent-child dyad. The domains can provide composite Total Stress score. An additional Life Stress domain assesses general stressful life events. High PSI scores indicate greater stress. The PSI demonstrates good validity and reliability psychometrics (Johnson, 2015).

Quality of Partner Relationship was measured at discharge and at follow-up via the Dyadic Adjustment Scale (DAS; (Spanier, 1989), a self-report measure that consists of 32 items that are summed for a Total Score; A higher score indicates higher quality relationship. The DAS has yielded a test-retest reliability above 0.90 and has been shown to successfully discriminate between distressed and non-distressed spouses in general and clinical samples.

Parental Social Support was assessed at discharge and at follow-up via the ENRICHED Social Support Index (ESSI; (Mitchell et al., 2003), a 7-item self-report questionnaire. Items are summed for a total score, with higher scores indicating greater social support. The ESSI has demonstrated strong psychometric properties in cardiac populations (Vaglio, et al., 2004).

Data Analysis

Comparisons were performed between parents who completed PTDS, and those who did not complete the PTDS to detect potential missingness bias. No statistically significant differences were observed on the study’s variables of interest, demonstrating no biased incompletion of the PTDS. Descriptive statistics were used to characterize demographic and clinical measures of parents and infants. Means, standard deviations, medians, and interquartile ranges were used to describe continuous variables. Frequencies and percentages were used to describe categorical variables. Generalized linear regression modeling was used to identify predictors of parental PTSS/PTSD at two time points (hospital discharge, and 4-month follow-up), for three main PTDS subscales: Total Symptom severity Score, Number of Symptoms, and PTSD diagnosis separately. First, potential (and time relevant) predictors were examined in separate bivariate models for each outcome at the two time points as follows: For outcomes at discharge we considered infant sex, birthweight, RACHS-1 score, cardiac physiology, timing of cardiac diagnosis, hospital length of stay, growth failure, feeding route, number of medications, parental age, race, ethnicity, parent education, income, insurance type, number of children, PSI Child Domain, PSI Parent Domain, PSI Life Stress, DAS, and ESSI; For outcomes at 4-month follow-up we considered the above (but PSI Child Domain, PSI Parent Domain, PSI Life Stress, DAS, and ESSI at 4-months measures) and also added number of readmissions and unscheduled cardiologist visits over the follow-up period. Next, all predictors significant at the α=0.20 level were entered to the initial multivariable models, and then we performed the backward stepwise process so that the final multivariable models included predictors significant at the α ≤ 0.05. Correlation matrices and variance inflation factors (VIFs) of final models were checked to detect multicollinearity (Cohen et al., 2013). Statistical analyses were performed using STATA 16™. The dataset is available upon request from the last author of the manuscript.

Results

The current analysis included 158 parents, of whom 97% were mothers. Majority of mothers were non-Hispanic (91%), white (80%), had collage education (66%), and belonged to a higher socioeconomic level (see Table 1). The infants were mostly diagnosed prenatally (82%), had lower RACHs-1 scores (62%) and were relatively split between one and two ventricle physiology (Table 1). Median hospital length of stay was 18 days (IRQ= 15). Among the 158 parents completing the PTDS at discharge, 14 (9%) met criteria for PTSD, half had more than five symptoms (IQR=7), and all had a mean total symptom severity score of 8.97 (SD=8.34). At four-month follow-up, 144 mothers completed the PTDS. Among these, twenty (14%) had a positive PTSD diagnosis, half had more than six symptoms (IQR=7), and all had a mean total symptom severity score of 8.96 (SD=8.29).

Table 1.

Demographic, illness related, and psychosocial factors of the study sample (N=158).

Demographic Factors	N (%)	Mean (SD)	Median (IQR)
Parental sex (female)	147 (96.7)
Parental Age		30.32 (5.33)	31 (8)
Parental Education (N = 139)
High School	23 (16.5)
College	116 (83.5)
Infant sex (female)	73 (46.8)
Race (N = 144)
White	115 (79.9)
Black	16 (11.1)
Other	13 (9.0)
Ethnicity
Hispanic	15 (9.1)
Non-Hispanic	143 (90.9)
Insurance Type
Private	115 (72.8)
Non-Private	43 (27.2)
Household Income (N= 140)
$0–24,999	23 (16.4)
$25,000–49,999	26 (18.6)
$50,000–99,999	47 (33.6)
>$100,000	44 (31.4)
Number of Children in Household		1.99 (1.03)	2 (2)
Illness-related Factors	N (%)	Mean (SD)	Median(IQR)
Single Ventricle Physiology	88 (55.7)
Prenatal Cardiac Diagnosis (Yes)	129 (81.6)
RACHs-1 score (N=117)
Low risk (categories 2–3)	73 (62.4)
High risk (categories 4–6)	44 (37.6)
Tube assisted Feeding at discharge	83 (52.5)
Gestational age (weeks)		38.83 (0.89)	39 (1)
Birthweight (g)		3333 (443.0)	3310 (657)
Number of Medications at discharge		2.85 (1.56)	3 (2)
Length of initial hospitalization (in days)		23.06 (14.78)	18 (15)
Number of Rehospitalizations		0.73 (1.05)	0 (1)
Emergency Department Visits		0.74 (1.10)	0 (1)
Unscheduled Cardiologist Visits		0.26 (0.62)	0 (1)
Psychosocial Factors	N (%)	Mean (SD)	Median(IQR)
Parenting stress, at discharge
Child domain (n=119)		104.59 (20.52)	103 (24)
Parents domain (n=141)		114.66 (26.61)	113 (34)
Life stress (n=152)		8.29 (6.30)	7 (7)
Dyadic Adjustment at discharge (n=149)		118.73 (15.52)	120 (18)
Perceived social support, at discharge		30.30 (4.12)	32 (5)
PTSS/PTSD outcomes (n=158)	N (%)	Mean (SD)	Median(IQR)
Parental PSTD at discharge
Meeting PTSD diagnostic criteria	14 (8.9%)
Total Symptom Severity		8.97 (8.34)	6 (10)
Number of Symptoms		6.45 (4.59)	5 (7)
Parental PSTD Scores at 4mo (n=144)
Meeting PTSD diagnostic criteria	20 (13.9%)
Total Symptom Severity		8.96 (8.29)	7 (9)
Number of Symptoms		6.53 (4.61)	6 (7)

Note. SD=Standard Deviation; IQR=Inter Quartile Range; RACHS-1= Risk Adjustment in Congenital Heart Surgery (Al-Radi et al., 2007); PTDS=Post-traumatic Stress Disorder Scale (Foa et al., 1997);

Bivariate modeling identified the following variables at the 0.20 significance level to be included in the multivariable analysis at discharge: weight-for-age Z score, number medications, length of hospital stay, feeding route, insurance type, parental education, DAS, ESSI, PSI Child and Parent Domains at the measured time point. At the four-month multivariable analysis the following were included: cardiac physiology, weight-for-age Z score, number medications, length of hospital stay, feeding route, number of readmissions, ED visits, unscheduled cardiologist visits, insurance type, parental education, parental age, ethnicity, number of children in the house, DAS, ESSI, PSI Child and Parent Domains, and Life Stress at the measured time point.

Final multivariable models following the stepwise procedure for parental PTSS/PTSD at discharge are presented in Table 2. Findings indicate that the PSI Parent Domain (β=0.17, p<0.001), parental college education (β =10.47, p<0.001), and number of medication (β =2.82, p=0.047) were all positively associated with the PTDS total severity scores at hospital discharge, and together explain 39% of the variance for this score. Similarly, PSI Parent Domain (β =0.08, p<0.001), parental college education (β =5.03, p<0.001), Medicaid insurance type (β = 7.78, p=0.004), and number of medication (β=2.42, p=0.001) were positively associated with the PTDS Number of Symptoms score, and the combination explained 42% of its variance. Lastly, tube assisted feeding route (OR=1.85, 95%CI=1.18–2.89) was the only remaining significant variable in the logistic regression model predicting PTSD according to diagnostic criteria (R²_pseudo =0.12).

Table 2.

Multivariable regression models for predictors of parental post-traumatic stress at discharge

	PTHS Total Severity subscale (n=107*)
Predictors	Estimate (SE)	95% CI	P-value	R2
PSI Parent Domain	0.17 (0.03)	0.12 0.22	<0.001	0.39
Parent education
High school	REF
College	10.47 (2.27)	5.45 15.50	<0.001
Number of medications	2.82 (1.40)	0.03 5.61	0.047
	PTDS Number of Symptoms subscale (n=107*)
	Estimate (SE)	95% CI	P-value	R2
PSI Parent Domain	0.08 (0.01)	0.05, 0.10	<0.001	0.42
Parent education
High school	REF	2.46, 7.57	<0.001
College	5.03 (1.44)	2.72, 7.33	<0.001
Insurance type
Private	REF
Medicaid	7.78 (2.64)	2.53, 13.02	0.004
Number of medications	2.42 (0.72)	0.99, 3.84	0.001
	PTSD Diagnosis subscale (n=76*)
	OR (SE)	95% CI	P-value	R2pseudo
Tube-assisted feeding	1.85 (0.42)	1.18, 2.89	0.007	0.120

Note. SE=Standard Error; CI =Confidence Intervals; PSI= Parenting Stress Index (Abidin, 1995); PTDS=Post-traumatic Stress Disorder Scale (Foa et el, 1997);

^*Sample sizes vary in the multivariable models, by the covariates’ missingness rates.

Table 3 presents the final multivariable models following the stepwise procedure for parental PTSS/PTSD at four-month follow-up. PSI Parent Domain (β=0.23, p<0.001), Hispanic Ethnicity (β=4.62, p=0.018), and ED visits (β = 1.32, p=0.031) were all associated with PTDS Total Severity score at four months, and explained 51% of the variance of this outcome. Similarly, PSI Parent Domain (β=0.12, p<0.001),Hispanic Ethnicity (β =3.45, p=0.001), and unscheduled cardiologist visits (β =1.03, p=0.049) were associated with PTDS Number of Symptoms score at four months, and explained 52% of the variance for this outcome. Finally, PSI Child (OR=1.07, 95%CI= 1.01–1.14) and Parent Domains (OR=1.07, 95%CI=1.02–1.14), single ventricle physiology (OR= 1.09, 95%CI=1.01–1.73), and number of children (OR=3.11, 95%CI=1.09–8.88) predicted positive PTSD diagnosis, with a R²_pseudo of 0.51 for this logistic model.

Table 3.

Multivariable regression models for predictors of parental post-traumatic stress at 4-month follow-up

	PTDS Total Severity subscale (n=106*)
	Estimate (SE)	95% CI	P-value	Model R2
PSI Parent Domain Ethnicity	0.23 (0.02)	0.18, 0.28	0.000	0.512
Non-Hispanic	REF
Hispanic	4.62 (1.91)	−0.82, 8.42	0.018
Number of ED visits	1.32 (0.60)	0.12 2.52	0.031
	PTDS Number of Symptoms subscale (n=106*)
	Estimate (SE)	95% CI	P-value	Model R2
PSI Parent Domain Ethnicity	0.12 (0.01)	0.09, 0.14	<0.001	0.52
Non-Hispanic	REF
Hispanic	3.45 (1.03)	1.39, 5.51	0.001
Number of Cardiologist visits	1.03 (0.51)	0.00, 2.06	0.049
	PTSD Diagnosis subscale (n = 101*)
	OR (SE)	95% CI	P-Value	R2pseudo
PSI Child Domain	1.07 (0.03)	1.01, 1.14	0.020	0.51
PSI Parent Domain	1.07 (0.03)	1.02, 1.14	0.009
Cardiac Physiology
Bi-ventricle	REF
Single ventricle	1.09 (1.10)	0.01, 1.73	0.024
Number of children	3.11 (1.66)	1.09, 8.88	0.033

Note. SE=Standard Error; CI =Confidence Intervals; PSI= Parenting Stress Index (Abidin, 1995); PTDS=Post-traumatic Stress Disorder Scale (Foa et al., 1997);

^*Sample sizes vary in the multivariable models, by the covariates’ missingness rates.

Discussion

According to our findings, 9% of mothers met the criteria for PTSD at the time of hospital discharge, and that number increased to 14% by the end of the study. Additionally a significant proportion of mothers had moderate PTSS at both time points. These results are consistent with prior studies and resemble the rates in the CHD and other pediatric populations (Helfricht et al., 2008; Landolt et al., 2003). Our results also indicate that parents may experience PTSS/PTSD over months following the CHD diagnosis and cardiac surgery, highlighting the importance of early screening while still in the hospital in order to provide proper support and interventions for these parents.

Further, we found that close to 40% of the early PTSS variance was explained by parental demographic and psychosocial characteristics (education level, insurance type, parental stress) and illness indicators (number of medications). Half of the variance of the PTSS at the end-of-study was similarly explained by parental characteristics (parental stress, ethnicity), and by illness indicators (ED and cardiologist visits). Early PTSD was predicted only by tube-assisted feeding at discharge, whereas PTSD at the study end was predicted by the combination of parental demographic and psychosocial characteristics (Parental stress on both Parent and Child domains, number of children, and cardiac physiology). While PTSS can develop immediately following the traumatic medical event (diagnosis/ surgery), PTSD can only be determined after a month of experienced symptoms. Therefore, it is potentially the end-of-study measurement that may better reflect the picture with regard to the PTSD predictors. As we also aware that parents of chronically ill children may be exposed to multiple traumatic experiences over the course of their child’s disease, we may assume that their post-traumatic stress is continuous and potential predictors may vary across time and illness contexts (Eagle & Kaminer, 2013; Kazak et al., 2006). Nevertheless, our findings may point towards parental psychosocial and behavioral factors, additionally to illness severity indicators, as potential risk factors to parental PTSS/PTSD in this population.

With regard to the demographic characteristics, ethnicity (Hispanic), SES (non-private insurance, more children in household), and education level (higher) were all positively associated with PTSS/PTSD. Higher rates of PTSS/PTSD have been previously demonstrated in lower SES communities or in minority ethnic groups (Muscara et al., 2015), and were attributed to the fewer available resources, or certain cultural illness beliefs affecting the coping and adaptation processes. Nevertheless, previous research has demonstrated mainly negative or no associations (Breslau et al., 1991; Franck et al., 2010; Nugent et al., 2007) between parental education level and PTSS/PTSD. Yet, some research suggests that higher education may lead to worse mental health outcomes, and vice versa. For example, Bonnano et al (Bonanno et al., 2007), found that education level was inversely associated with individual’s resilience after exposure to traumatic events, and Li and colleagues (Li et al., 2012) reported on a greater post-traumatic growth in moderately educated parents, compared to other education categories (low or high) in the CHD population. It is possible that highly educated parents may be more exposed to, or in the search for the publicly available information with regard to the potential risks and illness complications. Clearly, as we can only speculate in explaining this finding, further research is warranted.

With regard to the illness indicators, tube assisted feeding, single ventricle physiology, and number of ED and unscheduled outpatient visits were positively associated with PTSS and PTSD. These findings join few previous reports, suggesting that parents of more severely ill infants may potentially be at increased risk to suffer from post-traumatic stress (Woolf et al., 2016). However, the majority of studies indicate no associations between illness severity and PTSS/PTSD (Bronner et al., 2008; Helfricht et al., 2008; Landolt et al., 2002; McCarthy et al., 2012; Shaw et al., n.d.), and in others the psychosocial indicators present a risk for mental health problems even after adjusting (Franck et al., 2015)for the illness severity factors, suggesting that these go above and beyond the child’s medical condition (Brewin et al., 2000; Woolf et al., 2016).

With regard to the parental psychosocial characteristics our study found that parenting stress was positively linked to PTSS and PTSD at both time points, and in all but one of the outcomes (PTSD diagnosis at discharge), so it appeared to be a very important predictor. Specifically, the Parent Domain was found to have a more central role across the various models. The Parent Domain reflects parent’s feelings of wellness and mental health, attachment, competence, and perceived ability to fulfill their parental roles. The role of parental competencies and coping strategies with their child’s illness in determining their mental health has been previously demonstrated in the CHD population and other pediatric populations in the context of post-hospitalization traumatic stress (Farley et al., 2007; Franck et al., 2015; Stokes et al., 2020). Only late PTSD was predicted by the parenting stress Child Domain, which includes mainly a child’s temperament and behavioral aspects, and parental expectations from the child. Indeed, temperament and behavioral characteristics of the infant can unfold over time. Previous studies have pointed towards the fact that parents of infants with complex CHD may perceive their baby as difficult due to excessive crying and fussing (attributed to neurological and oxygenation issues) (Marino & Lipshitz, 1991; Torowicz et al., 2010). More focused attention is needed on parent mental health assessment to identify parents at risk. Additional research is needed to examine the role of the parental coping strategies and processes with accordance to the illness trajectory, to identify resilience factors among parents of infants with complex CHD. Early psychoeducational interventions during the hospitalization period can provide parents with adequate parenting coping skills, or other stress-reduction skills which may promote their well-being and perhaps reduce the risk for PTSS/PTSD (Golfenshtein et al., 2015; Kazak et al., 1997, 2005; Kazak, Alderfer, Streisand, et al., 2004).

Limitations

The secondary nature of the analysis limited our ability to select variables and measures for this study, and introduced additional limitations such as the fixed sample size. Future studies should also consider parental illness perceptions, coping strategies, and family functioning as potential contributors to the parental post-traumatic stress. Our sample of parents was relatively homogeneous, and included mainly mothers. Future studies should include both parents from more diverse populations from a wide range of race, ethnicity and SES backgrounds.

Conclusions

Findings point towards parental psychosocial and behavioral factors, additionally to illness and sociodemographic indicators, as potential risk factors to parental PTSS/PTSD in this population. More focused attention is needed on parent mental health assessment to identify parents at risk. Focused interventions to mitigate parenting stress may help to reduce the development of PTSS/PTSD.

Highlights.

• Our models explained 40–50% of the post-traumatic symptomology variance

• Demographics, psychosocial, and illness indicators predicted parental PTSS

• parents of sicker infants are at increased risk to experience PTSS and PTSD

• Parenting stress is a strong important predictor of PTSS and PTSD at both time points